Ubuntu 18.04 LTS optimised for security, multi-cloud, containers and AI

Kubeflow, the Google approach to TensorFlow on Kubernetes, and a range of CI/CD tools are integrated in Canonical Kubernetes and aligned with Google GKE for on-premise and on-cloud AI development.

"Having an OS that is tuned for advanced workloads such as AI and ML is critical to a high velocity team” said David Aronchick, Product Manager, Cloud AI at Google. “With the release of Ubuntu 18.04 LTS and Canonical's collaborations to the Kubeflow project, Canonical has provided both a familiar and highly performant operating system that works everywhere. Whether on-premise or in the cloud, software engineers and data scientists can use tools they are already familiar with, such as Ubuntu, Kubernetes and Kubeflow, and greatly accelerate their ability to deliver value for their customers."

Hardware acceleration with NVIDIA GPUs  is integrated in Ubuntu 18.04 LTS cloud images and Canonical’s OpenStack and Kubernetes distributions for on-premise bare metal operations, supporting Kubeflow and other machine learning / AI workflows.

Private cloud alternative to VMware

Canonical OpenStack delivers private cloud with significant savings over VMware and provides a modern, developer-friendly API. With built-in support for NFV and NVIDIA Tesla GPUs as well as other GPUs, the Canonical OpenStack offering has become a reference cloud for digital transformation workloads. Today, Ubuntu is at the heart of the world’s largest OpenStack clouds, both public and private, in key sectors such as finance, media, retail and telecommunications.

Kubernetes

Canonical’s Distribution of Kubernetes (CDK) runs on public clouds, VMware, OpenStack, and bare metal and delivers the latest upstream version, currently Kubernetes 1.10. After the initial three-step guided deployment, the distribution supports upgrades to future versions of Kubernetes, expansion of the Kubernetes cluster on demand, and integration with optional components for storage, networking and monitoring. A range of partners deliver their solutions on CDK, such as Rancher 2.0.

Platform for AI and machine learning

CDK supports GPU acceleration of workloads using the NVIDIA device plugin for Kubernetes. Complex workloads like Kubeflow that leverage NVIDIA GPUs ‘just work’ on CDK, reflecting joint efforts with Google to accelerate machine learning in the enterprise and providing a portable way to develop and deploy ML applications at scale. Applications built and tested with Kubeflow and CDK are perfectly transportable to Google Cloud.

Developers on Ubuntu can create applications on their workstations, test them on private bare-metal Kubernetes with CDK, and run them across vast data sets on Google’s GKE. The resulting models and inference engines can be delivered to Ubuntu devices at the edge of the network, creating a perfect pipeline for machine learning from workstation, to rack, to cloud and device.

Containers for legacy workloads with LXD 3.0
LXD 3.0 enables ‘lift-and-shift’ of legacy workloads into containers for performance and density, an essential part of the enterprise container strategy. While new applications will be built with containers and Kubernetes in mind, the long tail of legacy applications represents the most immediate benefit for data center operators interested in containers.

LXD provides ‘machine containers’, which behave like virtual machines in that they contain a full and mutable Linux guest operating system such as Ubuntu, RHEL or CentOS. That provides a traditional administration environment for legacy applications, which run at bare metal speeds with no hypervisor latency. Customers using unsupported or end-of-life Linux environments that have not received fixes for critical issues like Meltdown and Spectre can lift and shift those workloads into LXD on Ubuntu 18.04 LTS with all the latest kernel security  fixes.

Acceleration of snaps, deployed across desktop to the cloud

With more than 3,000 snaps published and millions installed, including official releases from Spotify, Skype, Slack and Firefox, snaps have become a popular way to get apps on Linux. Snaps are fully integrated in Ubuntu GNOME 18.04 LTS and KDE Neon. Publishers deliver updates directly and security is maintained with enhanced kernel isolation and system service mediation.

"Snaps provide a platform for us to deliver our CLI reliably” said Jeff Dickey, CLI Engineer, Heroku. “The automatic updates ensure our users are always using the latest version which saves us a big support headache. We're very happy to see snaps supported on more and more Linux distributions and becoming the standard for delivering software across Linux.”

Snaps work on desktops, devices, cloud virtual machines and bare-metal servers, providing a consistent delivery mechanism for applications and frameworks. Having an identical platform from workstation to edge and cloud accelerates global deployments and operations. Ubuntu 18.04 LTS features a default GNOME desktop in a family of desktop flavours including KDE, MATE and Budgie.

Ultra fast Ubuntu on your Windows desktop

New Hyper-V optimised images developed in collaboration with Microsoft enhance the virtual machine experience of Ubuntu in Windows. “In our upcoming OS release this spring, Hyper-V’s Quick Create VM Gallery will now include an image for the latest Ubuntu 18.04 LTS, officially stamped straight from Canonical” said Craig Wilhite, Program Manager, Microsoft. “This Ubuntu VM image will come pre-configured to offer clipboard functionality, drive redirection, dynamic resizing of VM console window, and much more as we look to provide a great Hyper-V client VM experience for Linux on Windows.”

Minimal desktop install

The popular new minimal desktop install provides only the core desktop and browser for those looking to save disk space and customise machines with their specific apps or requirements. In corporate environments, the minimal desktop serves as a base for custom desktop images, reducing the security cross-section of the platform.